AKI (Acute Kidney Injury) is a disease caused by a variety of etiological factors that results in a dramatic decline in kidney function over a short period of time, affecting approximately 13.3 million patients per year globally. The aim of this study was to construct a framework for predicting the risk of death in patients with AKI in the ICU (Intensive Care Unit) using the Bagging Ensemble Method. Using the MIMIC-IV database, the dataset was divided into a training set and a validation set in a ratio of 7:3. The Bagging Ensemble Method was employed and continuously optimized by Random Forest to predict the risk of death in patients. We compared the models with different structures (ROC: LR: 0.68, KNN: 0.75, RF0.80, CatBoost: 0.83, LightGBM: 0.85, Bagging: 0.91) and validated them against an independent dataset (ROC: Bagging: 0.87), as well as against traditional clinical scoring systems (SOFA: 0.61, APACHE-IV: 0.61). We also comprehensively evaluated the predictive performance of the model by precision, accuracy, recall, and F1-score, and the results showed that our proposed ensemble method has a good predictive effect. In addition, we conducted an in-depth analysis of the features affecting the model results using the SHAP model interpretable technique.